JPH01221957A - System for protecting bit error of packet sequence number - Google Patents

Abstract

PURPOSE:To perform the bit error protection of a packet sequence number with simple circuit constitution by containing a sequence number code having an inter-code distance with respect to the packet sequence number in the packet of a packet switching system, and deciding a packet number corresponding to the number code. CONSTITUTION:When a voice packet 60 with the sequence number 0 is transmitted from the transmission origin packet switching unit of a system, it is inputted from a packet input terminal A to a shift register 10 via the transmission path of a switching network, and the sequence number code 50 of the packet 60 is sent from a signal line 100 to a ROM12 for bit error correction. When the ROM12 inputs 000 of the code 50, the address of the ROM12 set in advance is referred, and ROM data 0 corresponding to the address is outputted from a signal line 102 as the sequence number of the packet 60. The sequence number is compared with the one sent from a (m) bit counter 16 at a comparator 18 inputting the sequence number 0, and 0 or 1 synchronized with the sequence number of the packet 60 is outputted. Thereby, it is possible to protect the bit error of the packet sequence number with the simple circuit constitution.

Description

DETAILED DESCRIPTION OF THE INVENTION (Field of Industrial Application) The present invention relates to a packet-switched system, and more particularly, to a bit error protection method for packet sequence numbers that is advantageously applied to voice packets transmitted in a packet-switched network. .

(Prior Art) Normally, error control is performed in a packet switching system, and retransmission processing is performed when an error occurs. However, since retransmission processing causes a delay in packets, this processing tends not to be performed when real-time performance is important, for example, in the case of voice packets. Therefore, when the sequence number of a voice packet becomes garbled during transmission, the data is corrected using a bit error protection scheme or nothing is done.

Fill:S (Fra
There is a meCheck 5equence) method. This is to provide an FCS check pit after the sequence number of a voice packet, and use this check bit to determine whether or not the sequence number has been transmitted correctly. If this method determines that the sequence number of the voice packet is incorrect, the correction circuit corrects the sequence number.

(Problems to be Solved by the Invention) This conventional method includes an FCS check circuit that checks the sequence number of a packet using an FCS check pit;
Since a correction circuit is required to correct the sequence number according to instructions from the FCS check circuit, there is a drawback that the circuit configuration to be realized becomes complex.

SUMMARY OF THE INVENTION An object of the present invention is to provide a bit error protection method for packet sequence numbers that can overcome the drawbacks of the prior art and can realize bit error protection with a simple circuit configuration.

(Means for Solving the Problems) In order to solve the above-mentioned problems, the present invention provides a bit error protection method for packet sequence numbers that corrects sequence numbers of packets whose data is garbled during transmission through a switching network of a packet switching system. , the packet includes a sequence number code having a predetermined inter-symbol distance related to the sequence number of the packet, and the packet switching system includes an extracting means for extracting the sequence number code of the packet from the packet, and a sequence number code of the packet according to the sequence number code of the packet. The bit error protection means determines the sequence number of the packet by removing a predetermined inter-symbol distance from the sequence number code extracted by the extraction means. Even if the sequence number code of a packet becomes garbled, the sequence number at the time of packet transmission is recovered.

(Function) According to the present invention, when a packet is input to the extraction means,
This extracts the packet's sequence number code and sends it to the bit error protection means. When the sequence number code is sent to the bit error protection means, the bit error protection means determines the sequence number corresponding to the input sequence number code and outputs it.

(Embodiment) Next, an embodiment of the packet sequence number bit error protection method according to the present invention will be described in detail with reference to the accompanying drawings.

FIG. 2 shows an example of the configuration of the audio bucket 80 in this embodiment. As shown in the figure, in the audio packet 80 of this embodiment, a sequence number code 50 indicating the sequence number of this packet is added before the information part 52 which is the audio signal. The sequence number code 50 consists of man bits, the m bits indicate the actual sequence number of the packet 60, and the n bits are used to detect bit errors in the actual sequence number indicated by the m bits.

FIG. 1 shows a functional block diagram of an error protection circuit of a packet switching system in which the packet sequence number bit error protection method according to the present invention is applied to voice packets. Note that components not directly related to the present invention, such as a packet disassembly/assembly device and a network switch, are not shown in the figure.

The shift register 10 is a register for extracting the sequence number code 50 from the voice packet 80. shift register l
Sequence number code 50 of audio packet 60 input to O
When shifted to a predetermined bit position according to the shift pulse, the bit error correction ROM ! is sent in parallel via the signal line 100. 2 is output.

The correction ROW 12 is a circuit that outputs the sequence number of the audio packet BO from the input sequence number code 50.

In other words, the correction ROM 12 is a man bit ROM.
It has addresses, and m-bit ROM data is allocated to each address. And correction RO
M12 is the input m◆n bit sequence number code 5
The ROM data is transferred from the storage location of the ROM address specified by 0 to the signal line 102 as the sequence number of the audio packet 80.
Output by Through this processing, even if the sequence number code 50 of the packet 80 becomes garbled during transmission, a correct sequence number is output.

The m-bit counter 16 is a counter that counts the sequence number of a series of packets 60 and outputs the result to the comparator 18 via an output 104. The comparator 18 is a comparator that receives the output 104 of the counter 16 and the output 102 of the correction ROW 12 via the signal line 10B, and compares these output values. In a packet-switched network, packets may be lost within the network due to, for example, a switch or transmission path failure, or packets may be discarded due to congestion within the network. Therefore, a series of packets 60
By counting the sequence number of the packet 60 and comparing it with the sequence number of the packet 60 actually sent by the comparator 18, it is determined whether the packet 60 has been lost or discarded.

The comparator 18 includes a counter 18 and a correction ROM 12.
If the values of the output packet sequence numbers are different, it is determined that the packet 60 has been lost or discarded within the packet network, and a loss detection signal is output.

FIG. 3 shows a correspondence diagram of the sequence number and sequence number code 50 of the packet 60 of the embodiment in which the sequence numbers are indicated by "0" and "1". When the sequence numbers are "0" and "1", 1 bit is sufficient as the number of bits m used for the sequence number. Further, when the number m of bits is 1 bit, the number n of bits used for bit error detection bits is twice that number, which is 2 bits. Therefore, the sequence number code 50 has m+nm3 bits. When the sequence number is "0", the two bit error detection bits are also "00" and the number code 50 is rooOJ. Also, when the sequence number is "1",
The bit error detection bit is “11” and the number code is 5.
0 becomes rill J.

FIG. 4 shows an example of a correspondence diagram of the ROM address of the bit error correction ROM 12 and the RO 14 data when the sequence number NEET 50 shown in FIG. 3 is used. In this example, as shown in Figure 4, the sequence number code is 5.
The ROM data is shown when the bit error protection number of 0 is set to 1 bit. That is, even if any one bit of the 3-bit sequence number code 50 becomes garbled during transmission, the sequence number of the packet 60 is set to the ROM data corresponding to the ROM address equal to the sequence number code. It can be modified to the actual order number at the time of shipment.

Specifically, the ROM address is "000", rooI
In the case of J, rollo J and "100", the ROM data becomes "0", and rollJ, rlolJ, r'1
When lOJ and rllllJ, the ROM data is rl
Become J. For this reason, the number code 50 is garbled, for example, from 1 bit "0" to "1" and becomes rool J, r.
For olo J and rlooJ, the ROM data becomes rQJ, similar to roooJ, and can be corrected to the actual sequence number at the time the packet 80 was sent.

Explain the operation. When the packet 60 with the sequence number rOJ is transmitted from the source packet switch (not shown) of the packet switching system, this packet 80 is sent from the packet input terminal A to the shift register 10 via the transmission path of the packet switching network. is input. When the shift register 10 receives the packet 60, the sequence number code 50 of the packet 130 is sent to the bit error correction RO via the signal line 100.
Send to M12.

At the time of transmission, the packet 60 with the sequence number rOJ has the sequence number code 50 set to rooo J and is sent from the source packet switch. Therefore, if the sequence number code 50 is not garbled during transmission, the error correction ROM
12 inputs the code 50 of rooOJ. When the error correction ROM 12 inputs the code 50 of rooOJ,
Refer to the preset ROM address, and
The ROM data corresponding to the ROM address of ooOJ, that is, "0", is output as the sequence number of the packet 80 via the signal line 102.

When the comparator 18 receives this sequence number rQJ via the signal line 106, it compares it with the sequence number sent from the m-bit counter 1B. In this embodiment, the sequence number of bucket) 80 is either "O" or "1". Therefore, in this embodiment, the counter 1B is rO
Either J or "1" is output to the comparator 18. Comparator 18 includes error correction law ROM 12 and counter 16
The input values are compared and a defect detection signal is output only if they are different. Therefore, even if packet 6o is lost or discarded during transmission, a loss detection signal is output, so that the packet switch can know that packet 60 is lost or discarded.

If one bit of the sequence number code 50 of the packet 60 becomes garbled during transmission, for example, when the LSB of code 50 becomes garbled and becomes rool J, the error correction ROM 12 will still
Since the code 50 of rool J is input and the ROM data of "0" corresponding to the ROM address of rool J is selected, the same sequence number as at the time of shipping can be output as described above.

When the voice packet 60 with the sequence number rlJ is transmitted next to the packet with the sequence number "0" from the source packet switch, only the sequence number code 50 is taken out by the shift register 10 in the same way as described above, and the sequence number code 50 is taken out from the error correction ROM. Sent to 12th. If the number code 50 is not garbled, the error correction ROM 12 receives the number need 50 of rlllJ and outputs the corresponding ROM data of rlJ as a sequence number. Also, if any one bit of sequence number 50 becomes garbled and becomes "0",
Since the ROM data is 1, the sequence number at the time of transmission can be output.

FIG. 6 shows an example of the structure of a conventional voice packet. As shown in the figure, the conventional audio packet 80 has a sequence number consisting of m bits at the beginning, followed by an FCS check pit consisting of 0 bits to check for garbled data in the sequence number. It is added before the information section, which is a signal. FIG. 5 shows a functional block diagram to which a bit error protection method according to the prior art is applied. As shown in the figure, in the conventional technology, bit error correction ROM1
This requires an FCS check circuit 20 for checking whether the sequence number of voice bucket 80 (instead of voice bucket 2) has been correctly transmitted using the FCS check pit, and a correction circuit 22 for correcting the sequence number according to instructions from the circuit 20.

Therefore, there was a problem that the circuit configuration became complicated.

On the other hand, in this embodiment, a code having an appropriate inter-symbol distance in the sequence number, that is, a sequence number code 50, is used, and the code 50 is used as the R of ROW 12 for bit error correction.
By making it correspond to the OM address, bit errors can be detected and corrected with a simple circuit configuration.

Note that in this embodiment, the sequence number bit is 1 bit, m=1
Although the case has been described, the present invention is of course not limited to this, and can be applied as long as m is a natural number. For example, when m=2, the bit error detection bit is n=4, and the sequence number code 50 is composed of 6 bits. Therefore, the ROM address is also 6 bits, and bit errors can be detected and corrected by using 64 types of ROM addresses. Similarly, when the sequence number code 50 is man bits, the ROM address also consists of man bits, and 2m+
Bit errors can be detected by using one type of ROM address,
Can be corrected. Further, in this embodiment, the present invention is advantageously applied to voice packets having a relatively small number of sequence numbers, but the present invention is of course applicable to data packets as well.

(Effects of the Invention) According to the present invention, a sequence number code having a sequence number code having an appropriate inter-symbol distance is used for the sequence number of the packet, and the sequence number of the packet is determined in accordance with this code. do. Therefore, bit error protection for packet sequence numbers can be performed using a circuit with a simple configuration.

[Brief explanation of the drawing]

FIG. 1 is a functional block diagram of an error protection circuit of a packet switch in which the packet sequence number bit error protection method according to the present invention is applied to voice packets. 2 is a configuration diagram of a voice packet in the protection system of the embodiment shown in FIG. 1, and FIG. 3 is a correspondence diagram showing an example of sequence numbers and corresponding sequence number codes in this embodiment. Fig. 4 is a data configuration diagram showing an example of correspondence between the ROM address and ROM data of the bit error correction ROM in the case of the sequence number code shown in Fig. 3, and Fig. 5 shows error protection of the packet switch according to the prior art Functional Block Diagram of Circuit FIG. 6 is a configuration diagram of a voice packet applied to the protection circuit of FIG. 5. Explanation of symbols of main parts 10, Shift register 12, ROM IE1 for bit error correction. ．． ．． m-bit counter 18, comparator 20, FCS check circuit 22, correction circuit 50, sequence number code 52, information section go? A1 Hi 4 Naru J Cat 2' Wake up to the audio bat 1 button? 11q Office Arrival Court H A] Hi 4th Row 31 Broadcasting Conventional Koto Bagut Section 84th Row 6 Days Hi-Nhi-fu W Around the Time GiiTi? o〆Tote Ichita's smile and A row No. 4-1

Claims (1)

[Claims] In a packet sequence number bit error protection method for correcting the sequence number of a packet whose data is garbled during transmission through a switching network of a packet switching system, the packet has a predetermined sequence number associated with the sequence number of the packet. The packet switching system includes an extraction means for extracting a sequence number code of the packet from the packet, and a bit error protection for determining the sequence number of the packet based on the sequence number code of the packet. means, the bit error protection means determines the order of the packet by removing the predetermined inter-symbol distance from the order number code extracted by the extraction means and determining the order number of the packet. A bit error protection method for packet sequence numbers, characterized in that even if the number code becomes garbled, the sequence number at the time of transmission of the packet is recovered.